Tachigali inca (Caesalpinioideae – Leguminosae), a new species of giant tree from Amazonian forests

The new species Tachigali inca is described and illustrated. It grows in lowland ‘terra firme’ forest of Amazonian Brazil, in the sub-Andean Amazon region of Peru, and in northeastern Bolivia. The species differs markedly from its most closely related species (T. amarumayu, T. prancei and T. setifera, all belonging to the “setifera group”), by the large cylindrical domatia on the leaf rachis, and by the brown-orange pulverulent indumentum on the abaxial surface of the leaflets and the young twigs; on older branches the indumentum becomes darker, degrades and then breaks off. It also differs by its linear petals with the upper half densely tomentose, the hairs forming small tufts.


INTRODUCTION
Tachigali Aubl. is a neotropical genus of leguminous trees widely distributed from the south of Mexico to southern Brazil and Bolivia. It is a member of subfamily Caesalpinioideae (LPWG 2017). Tachigali contains an estimated 75 species and is most speciose and morphologically diverse in the Amazon forest (Dwyer 1954(Dwyer , 1957a(Dwyer , 1957bvan der Werff 2008), with approximately 60 species (Huamantupa-Chuquimaco et al. 2019;van der Werff 2013).
Tachigali was first described by Aublet (1775), since then until recent years with several studies and others recently has been discovering up to an estimate of about 90 species, with more than 75% found in the Amazon region (Huamantupa-Chuquimaco et al. 2019;Huamantupa-Chuquimaco et al., unpubl. data). Tachigali is among the tree genera with the widest distri-bution in the Amazon region, for example T. paniculata and T. vaupesiana are recognized as hyperdominant tree species (ter Steege et al. 2013(ter Steege et al. , 2019. The Amazon region below the foothills of the eastern slopes of the Andes (Eva et al. 2005) is considered as a global biodiversity hotspot (Myers et al. 2000), and here the diversity of Tachigali is high. For example, the Amazonian forest of Cusco in southern Peru is home to 12 species of the genus (Huamantupa-Chuquimaco et al. 2016).
Tachigali also presents several specimens collected and monitored in permanent tree plots (Baker et al. 2014), however when they are collected without flowers or fruits they are difficult to identify, for example in RAINFOR plots in Peru they can only be determined up to a just over 50% (Baker et al. 2017). Recently, emphasis has been given to studying species complexes within Tachigali. One of these is the informal group "setifera" group comprising three species: T. amarumayu, T. prancei, and T. setifera, which has the shared morphological characteristics of the linear petals with an apical tuft of tomentose hairs and the abaxial surface of the leaflets with radially grouped hairs forming circles (Huamantupa-Chuquimaco et al. 2019). The group is supported in a preliminary molecular analysis which also reveals several undescribed species (Huamantupa-Chuquimaco et al., unpubl. data).
Here we describe a new species of Tachigali related to the "setifera group" from the Amazon forests of Bolivia, Brazil, and Peru.

MATERIALS AND METHODS
We conducted fieldwork at a number of localities in 'terra firme' forest of the Amazon region of Bolivia, Brazil, and Peru. In addition, we examined specimens in the following herbaria: Bolivia (LPB, USZ), Brazil (CEN, IAN, MG, RB, RON, UB), Peru (AMAZ, CUZ, MOL, SUMPA, USM), and the USA (F, MO, NY), (acronyms according to Thiers et al. 2019).

Diagnosis
Tachigali inca differs from the most closely related species T. amarumayu, T. prancei, and T. setifera by having trunk fenestrate (vs. not in the others three), brownorange pulverulent indumentum with some black dots, and in mature leaflets the indumentum having turned black on the abaxial surface of the leaflets (vs. the radially grouped hairs forming circles indumentum, on the abaxial surface of the leaflets), and the cylindrical, slightly ribbed laterally domatia immersed in the leaf rachis and on part of the petiole (vs. domatia absent).

Etymology
The epithet "inca" is named in honor to the Inca culture, who inhabited a large part of the Amazon at the base of the eastern Andes and part of central-southern Amazon, in which Tachigali inca is distributed.

Distribution and ecology
Tachigali inca is well distributed in Amazonian terra firme forest region, especially along the Andean foothills of Peru at a maximum of 700 meters elevation, and part of central-southern Amazon, into the north-west of Bolivia, and extending to the low Amazon of western Brazil in the states of Acre, Amazonas, Mato grosso and Rondônia (Fig. 3). Tachigali inca inhabits forests associated with clay soils and white sands.

Phenology
Collected in flower from January to May, in fruit from May to November. During flowering the entire tree crown is yellow and attracts many pollinators, including bees and butterflies.

Conservation status
Based on the extent of occurrence estimated at 1,024,042.723 km 2 , the Conservation status of T. inca is provisionally assessed as being of Least Concern (LC). However, in some areas such as central Amazonian of Peru, the white sand forests where it lives are threatened by the oil palm crops that are being implemented in recent years.

Common names and uses
Tachigali inca species are known in Brazil as 'tachi' or 'taxi' (meaning ant), as mentioned on the field label of Campbell 6419 collection. In Peru, common names of T. inca include 'tangarana', and 'palisanto' (Gutierrez 43) and 'ucshaquiro colorado' (Begazo 111). These three names refer to the presence of ants with strong formic acid that live in the domatias. The hard wood and straight trunk is ideal for local construction timber (Huamantupa-Chuquimaco et al. 2016).

Remarks
Tachigali inca differs from other species in the "setifera group" (T. amarumayu, T. prancei and T. setifera), mainly by its trunk being irregularly fenestrate; the abaxial surface of leaflets with a brown-orange pulverulent indumentum, which on old leaflets gradually degrades and detaches until the surface becomes a dark color (this characteristic has not been seen in any other species of the genus); and the cylindrical, slightly ribbed domatia immersed in the leaf rachis and on part of petiole (Fig. 2, Table 1), a type of domatia that is unknown in all other species of Tachigali. The other three species of the "setifera group" do not have a fenestrate trunk, their leaflets abaxial surfaces have sparsely or densely radiate clusters of hairs, with some additional free ferruginous hairs on the veins, and they all lack domatia. Tachigali amarumayu also has leaflets in 5-11 (often 8) pairs, and secondary veins on each leaflet in 11-23 pairs (Huamantupa-Chuquimaco et al. 2019). Tachigali prancei shares stamens and petals of a similar type with T. inca (table 1). Additional comparison between the four species is presented in Table 1.
Previously van der Werff (2008) considered some specimens of T. inca to belong within T. setifera sensu lato, by the similar leaflets shape and petals indumentum. Historically, other taxonomists have treated the majority of the specimens of T. inca as either T. amarumayu, T. setifera or T. vasquezii (Baker et al. 2017).
In a recent phylogenetic analysis, the "setifera group" is supported as distinct from all other Tachigali species and is geographically associated with species from the Amazon region of the eastern Andean foothills (Huamantupa-Chuquimaco et al., unpubl. data).